OSDN Git Service

git-core/git.git
6 years agoMerge branch 'js/win32-lazyload-dll'
Junio C Hamano [Thu, 28 Sep 2017 05:47:57 +0000 (14:47 +0900)]
Merge branch 'js/win32-lazyload-dll'

Add a helper in anticipation for its need in a future topic RSN.

* js/win32-lazyload-dll:
  Win32: simplify loading of DLL functions

6 years agoMerge branch 'jc/merge-x-theirs-docfix'
Junio C Hamano [Thu, 28 Sep 2017 05:47:57 +0000 (14:47 +0900)]
Merge branch 'jc/merge-x-theirs-docfix'

The documentation for '-X<option>' for merges was misleadingly
written to suggest that "-s theirs" exists, which is not the case.

* jc/merge-x-theirs-docfix:
  merge-strategies: avoid implying that "-s theirs" exists

6 years agoMerge branch 'ks/doc-use-camelcase-for-config-name'
Junio C Hamano [Thu, 28 Sep 2017 05:47:56 +0000 (14:47 +0900)]
Merge branch 'ks/doc-use-camelcase-for-config-name'

Doc update.

* ks/doc-use-camelcase-for-config-name:
  doc: camelCase the config variables to improve readability

6 years agoMerge branch 'mk/diff-delta-avoid-large-offset'
Junio C Hamano [Thu, 28 Sep 2017 05:47:56 +0000 (14:47 +0900)]
Merge branch 'mk/diff-delta-avoid-large-offset'

The delta format used in the packfile cannot reference data at
offset larger than what can be expressed in 4-byte, but the
generator for the data failed to make sure the offset does not
overflow.  This has been corrected.

* mk/diff-delta-avoid-large-offset:
  diff-delta: do not allow delta offset truncation

6 years agoMerge branch 'mk/diff-delta-uint-may-be-shorter-than-ulong'
Junio C Hamano [Thu, 28 Sep 2017 05:47:56 +0000 (14:47 +0900)]
Merge branch 'mk/diff-delta-uint-may-be-shorter-than-ulong'

The machinery to create xdelta used in pack files received the
sizes of the data in size_t, but lost the higher bits of them by
storing them in "unsigned int" during the computation, which is
fixed.

* mk/diff-delta-uint-may-be-shorter-than-ulong:
  diff-delta: fix encoding size that would not fit in "unsigned int"

6 years agoMerge branch 'rs/resolve-ref-optional-result'
Junio C Hamano [Thu, 28 Sep 2017 05:47:56 +0000 (14:47 +0900)]
Merge branch 'rs/resolve-ref-optional-result'

Code clean-up.

* rs/resolve-ref-optional-result:
  refs: pass NULL to resolve_ref_unsafe() if hash is not needed
  refs: pass NULL to refs_resolve_ref_unsafe() if hash is not needed
  refs: make sha1 output parameter of refs_resolve_ref_unsafe() optional

6 years agoMerge branch 'rs/mailinfo-qp-decode-fix'
Junio C Hamano [Thu, 28 Sep 2017 05:47:55 +0000 (14:47 +0900)]
Merge branch 'rs/mailinfo-qp-decode-fix'

"git mailinfo" was loose in decoding quoted printable and produced
garbage when the two letters after the equal sign are not
hexadecimal.  This has been fixed.

* rs/mailinfo-qp-decode-fix:
  mailinfo: don't decode invalid =XY quoted-printable sequences

6 years agoMerge branch 'jk/doc-read-tree-table-asciidoctor-fix'
Junio C Hamano [Thu, 28 Sep 2017 05:47:55 +0000 (14:47 +0900)]
Merge branch 'jk/doc-read-tree-table-asciidoctor-fix'

A docfix.

* jk/doc-read-tree-table-asciidoctor-fix:
  doc: put literal block delimiter around table

6 years agoMerge branch 'ik/userdiff-html-h-element-fix'
Junio C Hamano [Thu, 28 Sep 2017 05:47:54 +0000 (14:47 +0900)]
Merge branch 'ik/userdiff-html-h-element-fix'

The built-in pattern to detect the "function header" for HTML did
not match <H1>..<H6> elements without any attributes, which has
been fixed.

* ik/userdiff-html-h-element-fix:
  userdiff: fix HTML hunk header regexp

6 years agoMerge branch 'jk/fallthrough'
Junio C Hamano [Thu, 28 Sep 2017 05:47:53 +0000 (14:47 +0900)]
Merge branch 'jk/fallthrough'

Many codepaths have been updated to squelch -Wimplicit-fallthrough
warnings from Gcc 7 (which is a good code hygiene).

* jk/fallthrough:
  consistently use "fallthrough" comments in switches
  curl_trace(): eliminate switch fallthrough
  test-line-buffer: simplify command parsing

6 years agoMerge branch 'jk/diff-blob'
Junio C Hamano [Thu, 28 Sep 2017 05:47:53 +0000 (14:47 +0900)]
Merge branch 'jk/diff-blob'

"git cat-file --textconv" started segfaulting recently, which
has been corrected.

* jk/diff-blob:
  cat-file: handle NULL object_context.path

6 years agoMerge branch 'hn/typofix'
Junio C Hamano [Thu, 28 Sep 2017 05:47:52 +0000 (14:47 +0900)]
Merge branch 'hn/typofix'

* hn/typofix:
  submodule.h: typofix

6 years agoMerge branch 'ic/fix-filter-branch-to-handle-tag-without-tagger'
Junio C Hamano [Thu, 28 Sep 2017 05:47:52 +0000 (14:47 +0900)]
Merge branch 'ic/fix-filter-branch-to-handle-tag-without-tagger'

"git filter-branch" cannot reproduce a history with a tag without
the tagger field, which only ancient versions of Git allowed to be
created.  This has been corrected.

* ic/fix-filter-branch-to-handle-tag-without-tagger:
  filter-branch: use hash-object instead of mktag
  filter-branch: stash away ref map in a branch
  filter-branch: preserve and restore $GIT_AUTHOR_* and $GIT_COMMITTER_*
  filter-branch: reset $GIT_* before cleaning up

6 years agoMerge branch 'jk/describe-omit-some-refs'
Junio C Hamano [Thu, 28 Sep 2017 05:47:52 +0000 (14:47 +0900)]
Merge branch 'jk/describe-omit-some-refs'

"git describe --match" learned to take multiple patterns in v2.13
series, but the feature ignored the patterns after the first one
and did not work at all.  This has been fixed.

* jk/describe-omit-some-refs:
  describe: fix matching to actually match all patterns

6 years agoSync with 2.14.2
Junio C Hamano [Tue, 26 Sep 2017 05:15:55 +0000 (14:15 +0900)]
Sync with 2.14.2

* maint:
  Git 2.14.2
  Git 2.13.6
  Git 2.12.5
  Git 2.11.4
  Git 2.10.5
  cvsimport: shell-quote variable used in backticks
  archimport: use safe_pipe_capture for user input
  shell: drop git-cvsserver support by default
  cvsserver: use safe_pipe_capture for `constant commands` as well
  cvsserver: use safe_pipe_capture instead of backticks
  cvsserver: move safe_pipe_capture() to the main package

6 years agoWin32: simplify loading of DLL functions
Johannes Schindelin [Mon, 25 Sep 2017 16:06:24 +0000 (18:06 +0200)]
Win32: simplify loading of DLL functions

Dynamic loading of DLL functions is duplicated in several places in Git
for Windows' source code.

This patch adds a pair of macros to simplify the process: the
DECLARE_PROC_ADDR(<dll>, <return-type>, <function-name>,
...<function-parameter-types>...) macro to be used at the beginning of a
code block, and the INIT_PROC_ADDR(<function-name>) macro to call before
using the declared function. The return value of the INIT_PROC_ADDR()
call has to be checked; If it is NULL, the function was not found in the
specified DLL.

Example:

        DECLARE_PROC_ADDR(kernel32.dll, BOOL, CreateHardLinkW,
                          LPCWSTR, LPCWSTR, LPSECURITY_ATTRIBUTES);

        if (!INIT_PROC_ADDR(CreateHardLinkW))
                return error("Could not find CreateHardLinkW() function";

if (!CreateHardLinkW(source, target, NULL))
return error("could not create hardlink from %S to %S",
     source, target);
return 0;

Signed-off-by: Karsten Blees <blees@dcon.de>
Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
Reviewed-by: Jonathan Nieder <jrnieder@gmail.com>
Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
6 years agodoc: camelCase the config variables to improve readability
Kaartic Sivaraam [Sat, 23 Sep 2017 04:56:17 +0000 (10:26 +0530)]
doc: camelCase the config variables to improve readability

References to multi-word configuration variable names in our
documentation must consistently use camelCase to highlight where
the word boundaries are, even though these are treated case
insensitively.

Fix a few places that spell them in all lowercase, which makes
them harder to read.

Signed-off-by: Kaartic Sivaraam <kaarticsivaraam91196@gmail.com>
Reviewed-by: Jonathan Nieder <jrnieder@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
6 years agoThe ninth batch for 2.15
Junio C Hamano [Mon, 25 Sep 2017 06:28:27 +0000 (15:28 +0900)]
The ninth batch for 2.15

Signed-off-by: Junio C Hamano <gitster@pobox.com>
6 years agoMerge branch 'ks/test-readme-phrasofix'
Junio C Hamano [Mon, 25 Sep 2017 06:24:10 +0000 (15:24 +0900)]
Merge branch 'ks/test-readme-phrasofix'

Doc updates.

* ks/test-readme-phrasofix:
  t/README: fix typo and grammatically improve a sentence

6 years agoMerge branch 'ow/rev-parse-is-shallow-repo'
Junio C Hamano [Mon, 25 Sep 2017 06:24:10 +0000 (15:24 +0900)]
Merge branch 'ow/rev-parse-is-shallow-repo'

"git rev-parse" learned "--is-shallow-repository", that is to be
used in a way similar to existing "--is-bare-repository" and
friends.

* ow/rev-parse-is-shallow-repo:
  rev-parse: rev-parse: add --is-shallow-repository

6 years agoMerge branch 'rj/test-ulimit-on-windows'
Junio C Hamano [Mon, 25 Sep 2017 06:24:09 +0000 (15:24 +0900)]
Merge branch 'rj/test-ulimit-on-windows'

On Cygwin, "ulimit -s" does not report failure but it does not work
at all, which causes an unexpected success of some tests that
expect failures under a limited stack situation.  This has been
fixed.

* rj/test-ulimit-on-windows:
  t9010-*.sh: skip all tests if the PIPE prereq is missing
  test-lib: use more compact expression in PIPE prerequisite
  test-lib: don't use ulimit in test prerequisites on cygwin

6 years agoMerge branch 'jk/info-alternates-fix'
Junio C Hamano [Mon, 25 Sep 2017 06:24:09 +0000 (15:24 +0900)]
Merge branch 'jk/info-alternates-fix'

A regression fix for 2.11 that made the code to read the list of
alternate object stores overrun the end of the string.

* jk/info-alternates-fix:
  read_info_alternates: warn on non-trivial errors
  read_info_alternates: read contents into strbuf

6 years agoMerge branch 'mh/for-each-string-list-item-empty-fix'
Junio C Hamano [Mon, 25 Sep 2017 06:24:09 +0000 (15:24 +0900)]
Merge branch 'mh/for-each-string-list-item-empty-fix'

Code cmp.std.c nitpick.

* mh/for-each-string-list-item-empty-fix:
  for_each_string_list_item: avoid undefined behavior for empty list

6 years agoMerge branch 'tb/test-lint-echo-e'
Junio C Hamano [Mon, 25 Sep 2017 06:24:09 +0000 (15:24 +0900)]
Merge branch 'tb/test-lint-echo-e'

The test linter has been taught that we do not like "echo -e".

* tb/test-lint-echo-e:
  test-lint: echo -e (or -E) is not portable

6 years agoMerge branch 'jk/revision-remove-cmdline-pathspec'
Junio C Hamano [Mon, 25 Sep 2017 06:24:09 +0000 (15:24 +0900)]
Merge branch 'jk/revision-remove-cmdline-pathspec'

Code clean-up that also plugs memory leaks.

* jk/revision-remove-cmdline-pathspec:
  pathspec doc: parse_pathspec does not maintain references to args
  revision: replace "struct cmdline_pathspec" with argv_array

6 years agoMerge branch 'ls/travis-scriptify'
Junio C Hamano [Mon, 25 Sep 2017 06:24:09 +0000 (15:24 +0900)]
Merge branch 'ls/travis-scriptify'

The scripts to drive TravisCI has been reorganized and then an
optimization to avoid spending cycles on a branch whose tip is
tagged has been implemented.

* ls/travis-scriptify:
  travis-ci: fix "skip_branch_tip_with_tag()" string comparison
  travis: dedent a few scripts that are indented overly deeply
  travis-ci: skip a branch build if equal tag is present
  travis-ci: move Travis CI code into dedicated scripts

6 years agoMerge branch 'aw/gc-lockfile-fscanf-fix'
Junio C Hamano [Mon, 25 Sep 2017 06:24:08 +0000 (15:24 +0900)]
Merge branch 'aw/gc-lockfile-fscanf-fix'

"git gc" tries to avoid running two instances at the same time by
reading and writing pid/host from and to a lock file; it used to
use an incorrect fscanf() format when reading, which has been
corrected.

* aw/gc-lockfile-fscanf-fix:
  gc: call fscanf() with %<len>s, not %<len>c, when reading hostname

6 years agoMerge branch 'hv/mv-nested-submodules-test'
Junio C Hamano [Mon, 25 Sep 2017 06:24:08 +0000 (15:24 +0900)]
Merge branch 'hv/mv-nested-submodules-test'

A test to demonstrate "git mv" failing to adjust nested submodules
has been added.

* hv/mv-nested-submodules-test:
  add test for bug in git-mv for recursive submodules

6 years agoMerge branch 'bw/git-clang-format'
Junio C Hamano [Mon, 25 Sep 2017 06:24:07 +0000 (15:24 +0900)]
Merge branch 'bw/git-clang-format'

"make style" runs git-clang-format to help developers by pointing
out coding style issues.

* bw/git-clang-format:
  Makefile: add style build rule
  clang-format: outline the git project's coding style

6 years agoMerge branch 'nm/imap-send-with-curl'
Junio C Hamano [Mon, 25 Sep 2017 06:24:07 +0000 (15:24 +0900)]
Merge branch 'nm/imap-send-with-curl'

"git imap-send" has our own implementation of the protocol and also
can use more recent libCurl with the imap protocol support.  Update
the latter so that it can use the credential subsystem, and then
make it the default option to use, so that we can eventually
deprecate and remove the former.

* nm/imap-send-with-curl:
  imap-send: use curl by default when possible
  imap_send: setup_curl: retreive credentials if not set in config file
  imap-send: add wrapper to get server credentials if needed
  imap-send: return with error if curl failed

6 years agoMerge branch 'ks/commit-do-not-touch-cut-line'
Junio C Hamano [Mon, 25 Sep 2017 06:24:07 +0000 (15:24 +0900)]
Merge branch 'ks/commit-do-not-touch-cut-line'

The explanation of the cut-line in the commit log editor has been
slightly tweaked.

* ks/commit-do-not-touch-cut-line:
  commit-template: change a message to be more intuitive

6 years agoMerge branch 'tg/refs-allowed-flags'
Junio C Hamano [Mon, 25 Sep 2017 06:24:07 +0000 (15:24 +0900)]
Merge branch 'tg/refs-allowed-flags'

API error-proofing which happens to also squelch warnings from GCC.

* tg/refs-allowed-flags:
  refs: strip out not allowed flags from ref_transaction_update

6 years agoMerge branch 'rs/archive-excluded-directory'
Junio C Hamano [Mon, 25 Sep 2017 06:24:07 +0000 (15:24 +0900)]
Merge branch 'rs/archive-excluded-directory'

"git archive", especially when used with pathspec, stored an empty
directory in its output, even though Git itself never does so.
This has been fixed.

* rs/archive-excluded-directory:
  archive: don't add empty directories to archives

6 years agoMerge branch 'ks/help-alias-label'
Junio C Hamano [Mon, 25 Sep 2017 06:24:07 +0000 (15:24 +0900)]
Merge branch 'ks/help-alias-label'

"git help co" now says "co is aliased to ...", not "git co is".

* ks/help-alias-label:
  help: change a message to be more precise

6 years agoMerge branch 'jn/per-repo-object-store-fixes'
Junio C Hamano [Mon, 25 Sep 2017 06:24:06 +0000 (15:24 +0900)]
Merge branch 'jn/per-repo-object-store-fixes'

Step #0 of a planned & larger series to make the in-core object
store per in-core repository object.

* jn/per-repo-object-store-fixes:
  replace-objects: evaluate replacement refs without using the object store
  push, fetch: error out for submodule entries not pointing to commits
  pack: make packed_git_mru global a value instead of a pointer

6 years agoMerge branch 'jk/write-in-full-fix'
Junio C Hamano [Mon, 25 Sep 2017 06:24:06 +0000 (15:24 +0900)]
Merge branch 'jk/write-in-full-fix'

Many codepaths did not diagnose write failures correctly when disks
go full, due to their misuse of write_in_full() helper function,
which have been corrected.

* jk/write-in-full-fix:
  read_pack_header: handle signed/unsigned comparison in read result
  config: flip return value of store_write_*()
  notes-merge: use ssize_t for write_in_full() return value
  pkt-line: check write_in_full() errors against "< 0"
  convert less-trivial versions of "write_in_full() != len"
  avoid "write_in_full(fd, buf, len) != len" pattern
  get-tar-commit-id: check write_in_full() return against 0
  config: avoid "write_in_full(fd, buf, len) < len" pattern

6 years agoMerge branch 'ez/doc-duplicated-words-fix'
Junio C Hamano [Mon, 25 Sep 2017 06:24:06 +0000 (15:24 +0900)]
Merge branch 'ez/doc-duplicated-words-fix'

Typofix.

* ez/doc-duplicated-words-fix:
  doc: fix minor typos (extra/duplicated words)

6 years agoMerge branch 'kd/doc-for-each-ref'
Junio C Hamano [Mon, 25 Sep 2017 06:24:06 +0000 (15:24 +0900)]
Merge branch 'kd/doc-for-each-ref'

Doc update.

* kd/doc-for-each-ref:
  doc/for-each-ref: explicitly specify option names
  doc/for-each-ref: consistently use '=' to between argument names and values

6 years agoMerge branch 'cc/subprocess-handshake-missing-capabilities'
Junio C Hamano [Mon, 25 Sep 2017 06:24:06 +0000 (15:24 +0900)]
Merge branch 'cc/subprocess-handshake-missing-capabilities'

Finishing touches to a topic already in 'master'.

* cc/subprocess-handshake-missing-capabilities:
  subprocess: loudly die when subprocess asks for an unsupported capability

6 years agoMerge branch 'kw/write-index-reduce-alloc'
Junio C Hamano [Mon, 25 Sep 2017 06:24:05 +0000 (15:24 +0900)]
Merge branch 'kw/write-index-reduce-alloc'

A hotfix to a topic already in 'master'.

* kw/write-index-reduce-alloc:
  read-cache: fix index corruption with index v4
  Add t/helper/test-write-cache to .gitignore

6 years agoMerge branch 'mg/name-rev-tests-with-short-stack'
Junio C Hamano [Mon, 25 Sep 2017 06:24:05 +0000 (15:24 +0900)]
Merge branch 'mg/name-rev-tests-with-short-stack'

A handful of tests to demonstrates a recursive implementation of
"name-rev" hurts.

* mg/name-rev-tests-with-short-stack:
  t6120: test describe and name-rev with deep repos
  t6120: clean up state after breaking repo
  t6120: test name-rev --all and --stdin
  t7004: move limited stack prereq to test-lib

6 years agomerge-strategies: avoid implying that "-s theirs" exists
Junio C Hamano [Mon, 25 Sep 2017 05:34:23 +0000 (14:34 +0900)]
merge-strategies: avoid implying that "-s theirs" exists

The description of `-Xours` merge option has a parenthetical note
that tells the readers that it is very different from `-s ours`,
which is correct, but the description of `-Xtheirs` that follows it
carelessly says "this is the opposite of `ours`", giving a false
impression that the readers also need to be warned that it is very
different from `-s theirs`, which in reality does not even exist.

Clarify it a bit to avoid misleading readers.

Signed-off-by: Junio C Hamano <gitster@pobox.com>
6 years agomailinfo: don't decode invalid =XY quoted-printable sequences
René Scharfe [Sat, 23 Sep 2017 18:04:40 +0000 (20:04 +0200)]
mailinfo: don't decode invalid =XY quoted-printable sequences

Decode =XY in quoted-printable segments only if X and Y are hexadecimal
digits, otherwise just copy them.  That's at least better than
interpreting negative results from hexval() as a character.

Reported-by: Jeff King <peff@peff.net>
Signed-off-by: Rene Scharfe <l.s.r@web.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
6 years agorefs: pass NULL to resolve_ref_unsafe() if hash is not needed
René Scharfe [Sat, 23 Sep 2017 09:45:04 +0000 (11:45 +0200)]
refs: pass NULL to resolve_ref_unsafe() if hash is not needed

This allows us to get rid of some write-only variables, among them seven
SHA1 buffers.

Signed-off-by: Rene Scharfe <l.s.r@web.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
6 years agorefs: pass NULL to refs_resolve_ref_unsafe() if hash is not needed
René Scharfe [Sat, 23 Sep 2017 09:44:57 +0000 (11:44 +0200)]
refs: pass NULL to refs_resolve_ref_unsafe() if hash is not needed

This allows us to get rid of two write-only variables, one of them
being a SHA1 buffer.

Signed-off-by: Rene Scharfe <l.s.r@web.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
6 years agorefs: make sha1 output parameter of refs_resolve_ref_unsafe() optional
René Scharfe [Sat, 23 Sep 2017 09:41:45 +0000 (11:41 +0200)]
refs: make sha1 output parameter of refs_resolve_ref_unsafe() optional

Allow callers of refs_resolve_ref_unsafe() to pass NULL if they don't
need the resolved hash value.  We already allow the same for the flags
parameter.  This new leniency is inherited by the various wrappers like
resolve_ref_unsafe().

Signed-off-by: Rene Scharfe <l.s.r@web.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
6 years agouserdiff: fix HTML hunk header regexp
Ilya Kantor [Sat, 23 Sep 2017 07:53:47 +0000 (10:53 +0300)]
userdiff: fix HTML hunk header regexp

Current HTML header regexp doesn't match headers without attributes.

So it fails to match <h1>...</h1>, while <h1 class="smth">...</h1> matches.

Make attributes optional to fix this.  The regexp is still far from
perfect, but now it at least handles the common case.

Signed-off-by: Ilya Kantor <iliakan@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
6 years agodoc: put literal block delimiter around table
Jeff King [Sat, 23 Sep 2017 05:55:26 +0000 (01:55 -0400)]
doc: put literal block delimiter around table

The git-read-tree manpage has a table that is meant to
be shown with its spacing exactly as it is in the source. We
mark it as a "literal paragraph" by indenting each line by
at least one space. This renders OK with asciidoc for both
the HTML and manpage versions.

But there are two problems when we render it with
asciidoctor.

The first is that some lines mix tabs and spaces.  Even if
asciidoctor is correctly configured for 8-space tabs, it
seems to handle this case differently, soaking up some of
the initial literal-paragraph spaces and mis-aligning the
table text.

The second problem is that the table uses blank lines to
group rows. But as blank lines separate paragraphs in
asciidoc, this actually means that each chunk of the table
is rendered in its own pre-formatted <div> block. This
happens even with vanilla asciidoc, but there's no visible
result because the literal paragraphs aren't styled in any
special way. But with asciidoctor (or at least the styles
used on git-scm.com), literal paragraphs are styled with a
different background.  This breaks the table into a visually
distracting sequence of chunks.

We can fix both by adding a literal-paragraph block
delimiter. That turns the whole table into a single block
(for both implementations) and causes asciidoctor to render
the indentation as it is in the source.

Reported-at: https://github.com/git/git-scm.com/issues/1023
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
6 years agoGit 2.14.2 v2.14.2
Junio C Hamano [Fri, 22 Sep 2017 05:51:37 +0000 (14:51 +0900)]
Git 2.14.2

Signed-off-by: Junio C Hamano <gitster@pobox.com>
6 years agoSync with 2.13.6
Junio C Hamano [Fri, 22 Sep 2017 05:50:02 +0000 (14:50 +0900)]
Sync with 2.13.6

Signed-off-by: Junio C Hamano <gitster@pobox.com>
6 years agoGit 2.13.6 v2.13.6
Junio C Hamano [Fri, 22 Sep 2017 05:49:24 +0000 (14:49 +0900)]
Git 2.13.6

Signed-off-by: Junio C Hamano <gitster@pobox.com>
6 years agoSync with 2.12.5
Junio C Hamano [Fri, 22 Sep 2017 05:48:08 +0000 (14:48 +0900)]
Sync with 2.12.5

Signed-off-by: Junio C Hamano <gitster@pobox.com>
6 years agoGit 2.12.5 v2.12.5
Junio C Hamano [Fri, 22 Sep 2017 05:47:41 +0000 (14:47 +0900)]
Git 2.12.5

Signed-off-by: Junio C Hamano <gitster@pobox.com>
6 years agoSync with 2.11.4
Junio C Hamano [Fri, 22 Sep 2017 05:45:30 +0000 (14:45 +0900)]
Sync with 2.11.4

Signed-off-by: Junio C Hamano <gitster@pobox.com>
6 years agoGit 2.11.4 v2.11.4
Junio C Hamano [Fri, 22 Sep 2017 05:44:45 +0000 (14:44 +0900)]
Git 2.11.4

Signed-off-by: Junio C Hamano <gitster@pobox.com>
6 years agoSync with 2.10.5
Junio C Hamano [Fri, 22 Sep 2017 05:43:13 +0000 (14:43 +0900)]
Sync with 2.10.5

Signed-off-by: Junio C Hamano <gitster@pobox.com>
6 years agoGit 2.10.5 v2.10.5
Junio C Hamano [Fri, 22 Sep 2017 05:42:22 +0000 (14:42 +0900)]
Git 2.10.5

Signed-off-by: Junio C Hamano <gitster@pobox.com>
6 years agoMerge branch 'jk/safe-pipe-capture' into maint-2.10
Junio C Hamano [Fri, 22 Sep 2017 05:34:34 +0000 (14:34 +0900)]
Merge branch 'jk/safe-pipe-capture' into maint-2.10

6 years agoMerge branch 'jk/cvsimport-quoting' into maint-2.10
Junio C Hamano [Fri, 22 Sep 2017 05:34:34 +0000 (14:34 +0900)]
Merge branch 'jk/cvsimport-quoting' into maint-2.10

6 years agoMerge branch 'jc/cvsserver' into maint-2.10
Junio C Hamano [Fri, 22 Sep 2017 05:34:34 +0000 (14:34 +0900)]
Merge branch 'jc/cvsserver' into maint-2.10

6 years agoMerge branch 'jk/git-shell-drop-cvsserver' into maint-2.10
Junio C Hamano [Fri, 22 Sep 2017 05:34:34 +0000 (14:34 +0900)]
Merge branch 'jk/git-shell-drop-cvsserver' into maint-2.10

6 years agofilter-branch: use hash-object instead of mktag
Ian Campbell [Thu, 21 Sep 2017 07:49:32 +0000 (08:49 +0100)]
filter-branch: use hash-object instead of mktag

This allows us to recreate even historical tags which would now be consider
invalid, such as v2.6.12-rc2..v2.6.13-rc3 in the Linux kernel source tree which
lack the `tagger` header.

    $ git rev-parse v2.6.12-rc2
    9e734775f7c22d2f89943ad6c745571f1930105f
    $ git cat-file tag v2.6.12-rc2 | git mktag
    error: char76: could not find "tagger "
    fatal: invalid tag signature file
    $ git cat-file tag v2.6.12-rc2 | git hash-object -t tag -w --stdin
    9e734775f7c22d2f89943ad6c745571f1930105f

Signed-off-by: Ian Campbell <ijc@hellion.org.uk>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
6 years agofilter-branch: stash away ref map in a branch
Ian Campbell [Thu, 21 Sep 2017 07:49:31 +0000 (08:49 +0100)]
filter-branch: stash away ref map in a branch

With "--state-branch=<branchname>" option, the mapping from old object names
and filtered ones in ./map/ directory is stashed away in the object database,
and the one from the previous run is read to populate the ./map/ directory,
allowing for incremental updates of large trees.

Signed-off-by: Ian Campbell <ijc@hellion.org.uk>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
6 years agofilter-branch: preserve and restore $GIT_AUTHOR_* and $GIT_COMMITTER_*
Ian Campbell [Thu, 21 Sep 2017 07:49:30 +0000 (08:49 +0100)]
filter-branch: preserve and restore $GIT_AUTHOR_* and $GIT_COMMITTER_*

These are modified by set_ident() but a subsequent patch would like to operate
on their original values.

Signed-off-by: Ian Campbell <ijc@hellion.org.uk>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
6 years agofilter-branch: reset $GIT_* before cleaning up
Ian Campbell [Thu, 21 Sep 2017 07:49:29 +0000 (08:49 +0100)]
filter-branch: reset $GIT_* before cleaning up

This is pure code motion to enable a subsequent patch to add code which needs
to happen with the reset $GIT_* but before the temporary directory has been
cleaned up.

Signed-off-by: Ian Campbell <ijc@hellion.org.uk>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
6 years agoconsistently use "fallthrough" comments in switches
Jeff King [Thu, 21 Sep 2017 06:25:41 +0000 (02:25 -0400)]
consistently use "fallthrough" comments in switches

Gcc 7 adds -Wimplicit-fallthrough, which can warn when a
switch case falls through to the next case. The general idea
is that the compiler can't tell if this was intentional or
not, so you should annotate any intentional fall-throughs as
such, leaving it to complain about any unannotated ones.

There's a GNU __attribute__ which can be used for
annotation, but of course we'd have to #ifdef it away on
non-gcc compilers. Gcc will also recognize
specially-formatted comments, which matches our current
practice. Let's extend that practice to all of the
unannotated sites (which I did look over and verify that
they were behaving as intended).

Ideally in each case we'd actually give some reasons in the
comment about why we're falling through, or what we're
falling through to. And gcc does support that with
-Wimplicit-fallthrough=2, which relaxes the comment pattern
matching to anything that contains "fallthrough" (or a
variety of spelling variants). However, this isn't the
default for -Wimplicit-fallthrough, nor for -Wextra. In the
name of simplicity, it's probably better for us to support
the default level, which requires "fallthrough" to be the
only thing in the comment (modulo some window dressing like
"else" and some punctuation; see the gcc manual for the
complete set of patterns).

This patch suppresses all warnings due to
-Wimplicit-fallthrough. We might eventually want to add that
to the DEVELOPER Makefile knob, but we should probably wait
until gcc 7 is more widely adopted (since earlier versions
will complain about the unknown warning type).

Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
6 years agocurl_trace(): eliminate switch fallthrough
Jeff King [Thu, 21 Sep 2017 06:23:24 +0000 (02:23 -0400)]
curl_trace(): eliminate switch fallthrough

Our trace handler is called by curl with a curl_infotype
variable to interpret its data field. For most types we
print the data and then break out of the switch. But for
CURLINFO_TEXT, we print data and then fall through to the
"default" case, which does the exact same thing (nothing!)
that breaking out of the switch would.

This is probably a leftover from an early iteration of the
patch where the code after the switch _did_ do something
interesting that was unique to the non-text case arms.
But in its current form, this fallthrough is merely
confusing (and causes gcc's -Wimplicit-fallthrough to
complain).

Let's make CURLINFO_TEXT like the other case arms, and push
the default arm to the end where it's more obviously a
catch-all.

Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
6 years agotest-line-buffer: simplify command parsing
Jeff King [Thu, 21 Sep 2017 06:22:43 +0000 (02:22 -0400)]
test-line-buffer: simplify command parsing

The handle_command() function matches an incoming command
string with a sequence of starts_with() checks. But it also
surrounds these with a switch on the first character of the
command, which lets us jump to the right block of
starts_with() without going linearly through the list.

However, each case arm of the switch falls through to the
one below it. This is pointless (we know that a command
starting with 'b' does not need to check any of the commands
in the 'c' block), and it makes gcc's -Wimplicit-fallthrough
complain.

We could solve this by adding a break at the end of each
block. However, this optimization isn't helping anything.
Even if it does make matching faster (which is debatable),
this is code that is run only in the test suite, and each
run receives at most two of these "commands". We should
favor simplicity and readability over micro-optimizing.

Instead, let's drop the switch statement completely and
replace it with an if/else cascade.

Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
6 years agocat-file: handle NULL object_context.path
Jeff King [Thu, 21 Sep 2017 06:21:40 +0000 (02:21 -0400)]
cat-file: handle NULL object_context.path

Commit dc944b65f1 (get_sha1_with_context: dynamically
allocate oc->path, 2017-05-19) changed the rules that
callers must follow for seeing if we parsed a path in the
object name. The rules switched from "check if the oc.path
buffer is empty" to "check if the oc.path pointer is NULL".
But that commit forgot to update some sites in
cat_one_file(), meaning we might dereference a NULL pointer.

You can see this by making a path-aware request like
--textconv without specifying --path, and giving an object
name that doesn't have a path in it. Like:

  git cat-file --textconv HEAD

which will reliably segfault.

Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
6 years agosubmodule.h: typofix
Han-Wen Nienhuys [Thu, 21 Sep 2017 12:43:37 +0000 (14:43 +0200)]
submodule.h: typofix

Signed-off-by: Han-Wen Nienhuys <hanwen@google.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
6 years agotravis-ci: fix "skip_branch_tip_with_tag()" string comparison
Lars Schneider [Thu, 21 Sep 2017 20:48:30 +0000 (22:48 +0200)]
travis-ci: fix "skip_branch_tip_with_tag()" string comparison

09f5e97 ("travis-ci: skip a branch build if equal tag is present",
2017-09-17) introduced the "skip_branch_tip_with_tag" function with
a broken string comparison. Fix it!

Reported-by: SZEDER Gábor <szeder.dev@gmail.com>
Signed-off-by: Lars Schneider <larsxschneider@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
6 years agopathspec doc: parse_pathspec does not maintain references to args
Jonathan Nieder [Thu, 21 Sep 2017 04:41:12 +0000 (21:41 -0700)]
pathspec doc: parse_pathspec does not maintain references to args

The command line arguments passed to main() are valid for the life of
a program, but the same is not true for all other argv-style arrays
(e.g.  when a caller creates an argv_array).  Clarify that
parse_pathspec does not rely on the argv passed to it to remain valid.

This makes it easier to tell that callers like "git rev-list --stdin"
are safe and ensures that that is more likely to remain true as the
implementation of parse_pathspec evolves.

Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
6 years agoMerge branch 'jk/leak-checkers'
Junio C Hamano [Thu, 21 Sep 2017 04:38:37 +0000 (13:38 +0900)]
Merge branch 'jk/leak-checkers'

Many of our programs consider that it is OK to release dynamic
storage that is used throughout the life of the program by simply
exiting, but this makes it harder to leak detection tools to avoid
reporting false positives.  Plug many existing leaks and introduce
a mechanism for developers to mark that the region of memory
pointed by a pointer is not lost/leaking to help these tools.

* jk/leak-checkers:
  git-compat-util: make UNLEAK less error-prone

6 years agorevision: replace "struct cmdline_pathspec" with argv_array
Jeff King [Wed, 20 Sep 2017 20:36:59 +0000 (16:36 -0400)]
revision: replace "struct cmdline_pathspec" with argv_array

We assemble an array of strings in a custom struct,
NULL-terminate the result, and then pass it to
parse_pathspec().

But then we never free the array or the individual strings
(nor can we do the latter, as they are heap-allocated when
they come from stdin but not when they come from the
passed-in argv).

Let's swap this out for an argv_array. It does the same
thing with fewer lines of code, and it's safe to call
argv_array_clear() at the end to avoid a memory leak.

Reported-by: Martin Ågren <martin.agren@gmail.com>
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
6 years agotest-lint: echo -e (or -E) is not portable
Torsten Bögershausen [Sun, 17 Sep 2017 05:43:18 +0000 (07:43 +0200)]
test-lint: echo -e (or -E) is not portable

Some implementations of `echo` support the '-e' option to enable
backslash interpretation of the following string.
As an addition, they support '-E' to turn it off.

However, none of these are portable, POSIX doesn't even mention them,
and many implementations don't support them.

A check for '-n' is already done in check-non-portable-shell.pl,
extend it to cover '-n', '-e' or '-E'.

Signed-off-by: Torsten Bögershausen <tboegi@web.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
6 years agogit-compat-util: make UNLEAK less error-prone
Jonathan Tan [Tue, 19 Sep 2017 22:10:04 +0000 (15:10 -0700)]
git-compat-util: make UNLEAK less error-prone

Commit 0e5bba5 ("add UNLEAK annotation for reducing leak false
positives", 2017-09-08) introduced an UNLEAK macro to be used as
"UNLEAK(var);", but its existing definitions leave semicolons that act
as empty statements, which will lead to syntax errors, e.g.

if (condition)
UNLEAK(var);
else
something_else(var);

would be broken with two statements between if (condition) and else.
Lose the excess semicolon from the end of the macro replacement text.

Signed-off-by: Jonathan Tan <jonathantanmy@google.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
6 years agofor_each_string_list_item: avoid undefined behavior for empty list
Michael Haggerty [Wed, 20 Sep 2017 05:27:05 +0000 (22:27 -0700)]
for_each_string_list_item: avoid undefined behavior for empty list

If you pass a newly initialized or newly cleared `string_list` to
`for_each_string_list_item()`, then the latter does

    for (
            item = (list)->items; /* NULL */
            item < (list)->items + (list)->nr; /* NULL + 0 */
            ++item)

Even though this probably works almost everywhere, it is undefined
behavior, and it could plausibly cause highly-optimizing compilers to
misbehave.  C99 section 6.5.6 paragraph 8 explains:

    If both the pointer operand and the result point to elements
    of the same array object, or one past the last element of the
    array object, the evaluation shall not produce an overflow;
    otherwise, the behavior is undefined.

and (6.3.2.3.3) a null pointer does not point to anything.

Guard the loop with a NULL check to make the intent crystal clear to
even the most pedantic compiler.  A suitably clever compiler could let
the NULL check only run in the first iteration, but regardless, this
overhead is likely to be dwarfed by the work to be done on each item.

This problem was noticed by Coverity.

[jn: using a NULL check instead of a placeholder empty list;
 fleshed out the commit message based on mailing list discussion]

Signed-off-by: Michael Haggerty <mhagger@alum.mit.edu>
Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
6 years agoread_info_alternates: warn on non-trivial errors
Jeff King [Tue, 19 Sep 2017 19:41:10 +0000 (15:41 -0400)]
read_info_alternates: warn on non-trivial errors

When we fail to open $GIT_DIR/info/alternates, we silently
assume there are no alternates. This is the right thing to
do for ENOENT, but not for other errors.

A hard error is probably overkill here. If we fail to read
an alternates file then either we'll complete our operation
anyway, or we'll fail to find some needed object. Either
way, a warning is good idea. And we already have a helper
function to handle this pattern; let's just call
warn_on_fopen_error().

Note that technically the errno from strbuf_read_file()
might be from a read() error, not open(). But since read()
would never return ENOENT or ENOTDIR, and since it produces
a generic "unable to access" error, it's suitable for
handling errors from either.

Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
6 years agoMerge branch 'jk/info-alternates-fix-2.11' into jk/info-alternates-fix
Junio C Hamano [Wed, 20 Sep 2017 02:33:06 +0000 (11:33 +0900)]
Merge branch 'jk/info-alternates-fix-2.11' into jk/info-alternates-fix

* jk/info-alternates-fix-2.11:
  read_info_alternates: read contents into strbuf

6 years agoread_info_alternates: read contents into strbuf
Jeff King [Tue, 19 Sep 2017 19:41:07 +0000 (15:41 -0400)]
read_info_alternates: read contents into strbuf

This patch fixes a regression in v2.11.1 where we might read
past the end of an mmap'd buffer. It was introduced in
cf3c635210.

The link_alt_odb_entries() function has always taken a
ptr/len pair as input. Until cf3c635210 (alternates: accept
double-quoted paths, 2016-12-12), we made a copy of those
bytes in a string. But after that commit, we switched to
parsing the input left-to-right, and we ignore "len"
totally, instead reading until we hit a NUL.

This has mostly gone unnoticed for a few reasons:

  1. All but one caller passes a NUL-terminated string, with
     "len" pointing to the NUL.

  2. The remaining caller, read_info_alternates(), passes in
     an mmap'd file. Unless the file is an exact multiple of
     the page size, it will generally be followed by NUL
     padding to the end of the page, which just works.

The easiest way to demonstrate the problem is to build with:

  make SANITIZE=address NO_MMAP=Nope test

Any test which involves $GIT_DIR/info/alternates will fail,
as the mmap emulation (correctly) does not add an extra NUL,
and ASAN complains about reading past the end of the buffer.

One solution would be to teach link_alt_odb_entries() to
respect "len". But it's actually a bit tricky, since we
depend on unquote_c_style() under the hood, and it has no
ptr/len variant.

We could also just make a NUL-terminated copy of the input
bytes and operate on that. But since all but one caller
already is passing a string, instead let's just fix that
caller to provide NUL-terminated input in the first place,
by swapping out mmap for strbuf_read_file().

There's no advantage to using mmap on the alternates file.
It's not expected to be large (and anyway, we're copying its
contents into an in-memory linked list). Nor is using
git_open() buying us anything here, since we don't keep the
descriptor open for a long period of time.

Let's also drop the "len" parameter entirely from
link_alt_odb_entries(), since it's completely ignored. That
will avoid any new callers re-introducing a similar bug.

Reported-by: Michael Haggerty <mhagger@alum.mit.edu>
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
6 years agot9010-*.sh: skip all tests if the PIPE prereq is missing
Ramsay Jones [Sun, 17 Sep 2017 22:58:18 +0000 (23:58 +0100)]
t9010-*.sh: skip all tests if the PIPE prereq is missing

Every test in this file, except one, is marked with the PIPE prereq.
However, that lone test ('set up svn repo'), only performs some setup
work and checks whether the following test should be executed (by
setting an additional SVNREPO prerequisite). Since the following test
also requires the PIPE prerequisite, performing the setup test, when the
PIPE preequisite is missing, is simply wasted effort. Use the skip-all
test facility to skip all tests when the PIPE prerequisite is missing.

Signed-off-by: Ramsay Jones <ramsay@ramsayjones.plus.com>
Reviewed-by: Jonathan Nieder <jrnieder@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
6 years agotest-lib: use more compact expression in PIPE prerequisite
Ramsay Jones [Sun, 17 Sep 2017 22:56:00 +0000 (23:56 +0100)]
test-lib: use more compact expression in PIPE prerequisite

Signed-off-by: Ramsay Jones <ramsay@ramsayjones.plus.com>
Reviewed-by: Jonathan Nieder <jrnieder@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
6 years agorev-parse: rev-parse: add --is-shallow-repository
Øystein Walle [Mon, 18 Sep 2017 17:04:29 +0000 (19:04 +0200)]
rev-parse: rev-parse: add --is-shallow-repository

Running `git fetch --unshallow` on a repo that is not in fact shallow
produces a fatal error message. Add a helper to rev-parse that scripters
can use to determine whether a repo is shallow or not.

Signed-off-by: Øystein Walle <oystwa@gmail.com>
Reviewed-by: Jonathan Nieder <jrnieder@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
6 years agot/README: fix typo and grammatically improve a sentence
Kaartic Sivaraam [Sun, 17 Sep 2017 10:18:15 +0000 (10:18 +0000)]
t/README: fix typo and grammatically improve a sentence

Signed-off-by: Kaartic Sivaraam <kaarticsivaraam91196@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
6 years agoThe eighth batch for 2.15
Junio C Hamano [Tue, 19 Sep 2017 01:55:19 +0000 (10:55 +0900)]
The eighth batch for 2.15

Signed-off-by: Junio C Hamano <gitster@pobox.com>
6 years agoMerge branch 'rk/commit-tree-make-F-verbatim'
Junio C Hamano [Tue, 19 Sep 2017 01:47:57 +0000 (10:47 +0900)]
Merge branch 'rk/commit-tree-make-F-verbatim'

Unlike "git commit-tree < file", "git commit-tree -F file" did not
pass the contents of the file verbatim and instead completed an
incomplete line at the end, if exists.  The latter has been updated
to match the behaviour of the former.

* rk/commit-tree-make-F-verbatim:
  commit-tree: do not complete line in -F input

6 years agoMerge branch 'rs/strbuf-leakfix'
Junio C Hamano [Tue, 19 Sep 2017 01:47:57 +0000 (10:47 +0900)]
Merge branch 'rs/strbuf-leakfix'

Many leaks of strbuf have been fixed.

* rs/strbuf-leakfix: (34 commits)
  wt-status: release strbuf after use in wt_longstatus_print_tracking()
  wt-status: release strbuf after use in read_rebase_todolist()
  vcs-svn: release strbuf after use in end_revision()
  utf8: release strbuf on error return in strbuf_utf8_replace()
  userdiff: release strbuf after use in userdiff_get_textconv()
  transport-helper: release strbuf after use in process_connect_service()
  sequencer: release strbuf after use in save_head()
  shortlog: release strbuf after use in insert_one_record()
  sha1_file: release strbuf on error return in index_path()
  send-pack: release strbuf on error return in send_pack()
  remote: release strbuf after use in set_url()
  remote: release strbuf after use in migrate_file()
  remote: release strbuf after use in read_remote_branches()
  refs: release strbuf on error return in write_pseudoref()
  notes: release strbuf after use in notes_copy_from_stdin()
  merge: release strbuf after use in write_merge_heads()
  merge: release strbuf after use in save_state()
  mailinfo: release strbuf on error return in handle_boundary()
  mailinfo: release strbuf after use in handle_from()
  help: release strbuf on error return in exec_woman_emacs()
  ...

6 years agoMerge branch 'jk/shortlog-ident-cleanup'
Junio C Hamano [Tue, 19 Sep 2017 01:47:56 +0000 (10:47 +0900)]
Merge branch 'jk/shortlog-ident-cleanup'

Code clean-up.

* jk/shortlog-ident-cleanup:
  shortlog: skip format/parse roundtrip for internal traversal

6 years agoMerge branch 'mh/packed-ref-transactions'
Junio C Hamano [Tue, 19 Sep 2017 01:47:56 +0000 (10:47 +0900)]
Merge branch 'mh/packed-ref-transactions'

Implement transactional update to the packed-ref representation of
references.

* mh/packed-ref-transactions:
  files_transaction_finish(): delete reflogs before references
  packed-backend: rip out some now-unused code
  files_ref_store: use a transaction to update packed refs
  t1404: demonstrate two problems with reference transactions
  files_initial_transaction_commit(): use a transaction for packed refs
  prune_refs(): also free the linked list
  files_pack_refs(): use a reference transaction to write packed refs
  packed_delete_refs(): implement method
  packed_ref_store: implement reference transactions
  struct ref_transaction: add a place for backends to store data
  packed-backend: don't adjust the reference count on lock/unlock

6 years agoMerge branch 'kw/merge-recursive-cleanup'
Junio C Hamano [Tue, 19 Sep 2017 01:47:56 +0000 (10:47 +0900)]
Merge branch 'kw/merge-recursive-cleanup'

A leakfix and code clean-up.

* kw/merge-recursive-cleanup:
  merge-recursive: change current file dir string_lists to hashmap
  merge-recursive: remove return value from get_files_dirs
  merge-recursive: fix memory leak

6 years agoMerge branch 'sb/merge-commit-msg-hook'
Junio C Hamano [Tue, 19 Sep 2017 01:47:56 +0000 (10:47 +0900)]
Merge branch 'sb/merge-commit-msg-hook'

As "git commit" to conclude a conflicted "git merge" honors the
commit-msg hook, "git merge" that recoreds a merge commit that
cleanly auto-merges should, but it didn't.

* sb/merge-commit-msg-hook:
  builtin/merge: honor commit-msg hook for merges

6 years agoMerge branch 'jk/leak-checkers'
Junio C Hamano [Tue, 19 Sep 2017 01:47:55 +0000 (10:47 +0900)]
Merge branch 'jk/leak-checkers'

Many of our programs consider that it is OK to release dynamic
storage that is used throughout the life of the program by simply
exiting, but this makes it harder to leak detection tools to avoid
reporting false positives.  Plug many existing leaks and introduce
a mechanism for developers to mark that the region of memory
pointed by a pointer is not lost/leaking to help these tools.

* jk/leak-checkers:
  add UNLEAK annotation for reducing leak false positives
  set_git_dir: handle feeding gitdir to itself
  repository: free fields before overwriting them
  reset: free allocated tree buffers
  reset: make tree counting less confusing
  config: plug user_config leak
  update-index: fix cache entry leak in add_one_file()
  add: free leaked pathspec after add_files_to_cache()
  test-lib: set LSAN_OPTIONS to abort by default
  test-lib: --valgrind should not override --verbose-log

6 years agoMerge branch 'nm/pull-submodule-recurse-config'
Junio C Hamano [Tue, 19 Sep 2017 01:47:55 +0000 (10:47 +0900)]
Merge branch 'nm/pull-submodule-recurse-config'

"git -c submodule.recurse=yes pull" did not work as if the
"--recurse-submodules" option was given from the command line.
This has been corrected.

* nm/pull-submodule-recurse-config:
  pull: honor submodule.recurse config option
  pull: fix cli and config option parsing order

6 years agoMerge branch 'mh/packed-ref-store-prep'
Junio C Hamano [Tue, 19 Sep 2017 01:47:55 +0000 (10:47 +0900)]
Merge branch 'mh/packed-ref-store-prep'

Fix regression to "gitk --bisect" by a recent update.

* mh/packed-ref-store-prep:
  rev-parse: don't trim bisect refnames

6 years agoMerge branch 'ma/remove-config-maybe-bool'
Junio C Hamano [Tue, 19 Sep 2017 01:47:55 +0000 (10:47 +0900)]
Merge branch 'ma/remove-config-maybe-bool'

Finishing touches to a recent topic.

* ma/remove-config-maybe-bool:
  config: remove git_config_maybe_bool

6 years agoMerge branch 'jk/system-path-cleanup'
Junio C Hamano [Tue, 19 Sep 2017 01:47:55 +0000 (10:47 +0900)]
Merge branch 'jk/system-path-cleanup'

Code clean-up.

* jk/system-path-cleanup:
  git_extract_argv0_path: do nothing without RUNTIME_PREFIX
  system_path: move RUNTIME_PREFIX to a sub-function

6 years agoMerge branch 'jh/hashmap-disable-counting'
Junio C Hamano [Tue, 19 Sep 2017 01:47:54 +0000 (10:47 +0900)]
Merge branch 'jh/hashmap-disable-counting'

Our hashmap implementation in hashmap.[ch] is not thread-safe when
adding a new item needs to expand the hashtable by rehashing; add
an API to disable the automatic rehashing to work it around.

* jh/hashmap-disable-counting:
  hashmap: add API to disable item counting when threaded

6 years agoMerge branch 'bb/doc-eol-dirty'
Junio C Hamano [Tue, 19 Sep 2017 01:47:54 +0000 (10:47 +0900)]
Merge branch 'bb/doc-eol-dirty'

Doc update.

* bb/doc-eol-dirty:
  Documentation: mention that `eol` can change the dirty status of paths

6 years agoMerge branch 'jt/packmigrate'
Junio C Hamano [Tue, 19 Sep 2017 01:47:53 +0000 (10:47 +0900)]
Merge branch 'jt/packmigrate'

Remove unneeded file added by a topic already in 'master'.

* jt/packmigrate:
  Remove inadvertently added outgoing/packfile.h

6 years agoMerge branch 'jk/incore-lockfile-removal'
Junio C Hamano [Tue, 19 Sep 2017 01:47:53 +0000 (10:47 +0900)]
Merge branch 'jk/incore-lockfile-removal'

The long-standing rule that an in-core lockfile instance, once it
is used, must not be freed, has been lifted and the lockfile and
tempfile APIs have been updated to reduce the chance of programming
errors.

* jk/incore-lockfile-removal:
  stop leaking lock structs in some simple cases
  ref_lock: stop leaking lock_files
  lockfile: update lifetime requirements in documentation
  tempfile: auto-allocate tempfiles on heap
  tempfile: remove deactivated list entries
  tempfile: use list.h for linked list
  tempfile: release deactivated strbufs instead of resetting
  tempfile: robustify cleanup handler
  tempfile: factor out deactivation
  tempfile: factor out activation
  tempfile: replace die("BUG") with BUG()
  tempfile: handle NULL tempfile pointers gracefully
  tempfile: prefer is_tempfile_active to bare access
  lockfile: do not rollback lock on failed close
  tempfile: do not delete tempfile on failed close
  always check return value of close_tempfile
  verify_signed_buffer: prefer close_tempfile() to close()
  setup_temporary_shallow: move tempfile struct into function
  setup_temporary_shallow: avoid using inactive tempfile
  write_index_as_tree: cleanup tempfile on error